Resource for a list of Solar Fires?

[Anode]

Does anyone have a resource for a list of fires that were caused by photovoltaic systems, or fires that were not caused by them, but had to be dealt with by first responders? A fire department engineer has expressed that there have been many fires, and this is a real issue.

The FD is concerned with clearance of solar to roof in a ballasted system, and a host of other issues some modeled after California's standards. If this set of criteria is not met, they are saying they will enforce a requirement for sprinklers to be installed between the solar modules and the roof deck. One of the main requirements is that the low edge of the solar module cannot be less than 9" from the roof deck, which categorically excludes many ballasted racking systems.

I have never heard of such a thing... I know this is not necessarily electrical code related, but definitely code related for solar. When asked about these codes, I got "we are sort of making this up as we go along". Which doesn't help my pre-planning phase.

Ultimately, I cannot blame them, I believe in increased safety, and some of you will disagree with me, but the intent of 690.12 as an example is a good thing in my opinion. Of course, I think having sprinklers installed for solar systems is perhaps going to an extreme length. Thoughts? Wisdom? Anyone have this magic list I am curious about?

 

[jaggedben]

The only fire I'm aware of is the Bakersfield fire, about which at least one article has been written chiefly concerning ground-fault protection. To my knowledge there have not been 'many fires'. My response to hearing someone saying that would be to throw it back at them. 'That's interesting, I haven't heard about many fires. Can you share some info about that so I can learn more?'

Racking systems do get subjected to fire tests and so I would look for a racking system that is Class A rated and tell them 'this is the highest industry standard, in my opinion you should not be asking for more than this.' It may be true that many ballasted systems aren't class A rated but that 9" rule is totally arbitrary as far as I know.

Hope that helps.

 

[pv_n00b]

There is no list that I know of. Occasionally I do a news search and something turns up, and over the years that has added up to many fires. They are almost never included in the PV industry news. The problem is that you just end up with a list of buildings with fires that just happen to have PV systems on them. Typical new coverage tends to be inaccurate on the technical details, they may say the PV system had something to do with the fire when it did not and they often misquote people on details.

The Bakersfield fire was very unusual in the amount of detail that came out to the PV industry. Ask anybody who does forensics work why details almost never get out and they will tell you it's the NDAs. I've done forensics and the NDAs prevent any promulgation of details because everybody involved is afraid of either being sued, being exposed as at fault, or generally just getting bad press. It's too bad because there are a lot of good lessons to be learned from the failures. People learned a lot about safe PV installation from the Bakersfield fire.

Put this search in Google News and read up: (photovoltaic OR solar) fire

The image search is particularly interesting. Lots of burned arrays on roofs.

 

[pv_n00b]

As for the OPs issue. Fire departments have been provided with lots of guidance at this point on PV array installation standards related to fire prevention and fighting fires, but if a local fire department chooses to ignore that guidance and make up their own there is not much you can do other than talk to them and try to get them the accept the guidance. You can also try contacting Matthew Paiss with the San Jose fire department. He does a lot of fire department training related to PV arrays and he might have some suggestions.

 

[Carultch]

There is no list that I know of. Occasionally I do a news search and something turns up, and over the years that has added up to many fires. They are almost never included in the PV industry news. The problem is that you just end up with a list of buildings with fires that just happen to have PV systems on them. Typical new coverage tends to be inaccurate on the technical details, they may say the PV system had something to do with the fire when it did not and they often misquote people on details.

The Bakersfield fire was very unusual in the amount of detail that came out to the PV industry. Ask anybody who does forensics work why details almost never get out and they will tell you it's the NDAs. I've done forensics and the NDAs prevent any promulgation of details because everybody involved is afraid of either being sued, being exposed as at fault, or generally just getting bad press. It's too bad because there are a lot of good lessons to be learned from the failures. People learned a lot about safe PV installation from the Bakersfield fire.

Put this search in Google News and read up: (photovoltaic OR solar) fire

The image search is particularly interesting. Lots of burned arrays on roofs.

So in otherwords, they can make details public if it were arson, but not if it were an honest-to-god accident or oversight. Is that correct?

 

[pv_n00b]

So in otherwords, they can make details public if it were arson, but not if it were an honest-to-god accident or oversight. Is that correct?

I don't know of any requirement to make the information public if there was a crime. NDA's typically say information can only be released to a 3rd party if demanded through legal means, like a court order. Even then it is released to the 3rd party and it may or may not be released to the public.

I don't remember why there was so much information released on the Bakersfield fire. I'll have to ask Bill sometime.

 

[pv_n00b]

There are several groups involved in forensic inspections of fires, insurance companies, fire departments, AHJs, 3rd party engineers, etc. Some of these groups are under NDAs and some are not. For instance if the AHJ or the fire department does an inspection and publishes the results then the people who are under NDA can talk about those published results because they are already public knowledge. But if the groups not under NDA never publish anything then the people under NDA can't release any information.

Both the Bakersfield and the Mount Holly fires had information released by groups not under NDA so that allowed others under NDA to write articles and do presentations. I hope this helps people understand why we hear so much about some fires and little about others. There are plenty of fires on structures with PV but not all make it to the news outside the local area, this is particularly true of residential fires.

 

[Anode]

You know, a better question for this thread would have been, has anyone ever heard of an AHJ requiring sprinklers beneath a pv array on a flat roof with a ballasted system.

What is worst is that they are not being totally forthcoming with the requirements, or with the codes that are deriving them. :? :happysad:

 

[pv_n00b]

I can't think of any reason for this other than, "we just want you to do it."

 

[jaggedben]

You know, a better question for this thread would have been, has anyone ever heard of an AHJ requiring sprinklers beneath a pv array on a flat roof with a ballasted system.

No. Hope that helps.

 

[SolarPro]

The data in the attached table are pretty interesting. Basically, SunPower—one of the largest commercial project developers in the world—found that once they closed the ground-fault detection blind spot, their PV systems stopped developing fires. They published these results here: "Resolving Fire Hazards From the Ground-Fault Detection Blind Spot."

In a related article, "The Heat Is On: Fault Detection and Fire Prevention," Bill Brooks has made a very compelling case that the industry should deploy PV systems built to NEC 2014 standards, which address both arc-fault detection and the ground-fault detection blind spot, regardless of what Code edition your AHJ enforces. This is a much more effective way to eliminate fires originating within a PV system than installing sprinkler systems under ballasted arrays.

 

[ggunn]

The data in the attached table are pretty interesting. Basically, SunPower—one of the largest commercial project developers in the world—found that once they closed the ground-fault detection blind spot, their PV systems stopped developing fires. They published these results here: "Resolving Fire Hazards From the Ground-Fault Detection Blind Spot."

In a related article, "The Heat Is On: Fault Detection and Fire Prevention," Bill Brooks has made a very compelling case that the industry should deploy PV systems built to NEC 2014 standards, which address both arc-fault detection and the ground-fault detection blind spot, regardless of what Code edition your AHJ enforces. This is a much more effective way to eliminate fires originating within a PV system than installing sprinkler systems under ballasted arrays.

And of course there is the point that throwing water on an electrical fire is not advisable.

 

[don_resqcapt19]

The fire departments concerns go much beyond the possibility of the solar equipment actually starting a fire. One of those issues is the panels limit the roof space that can be used for cutting ventilation holes and they limit movement on the roof. The issue of water on energized electrical systems is also a concern, but by careful selection of nozzel pattern and pressure you can safely put water on energized electrical systems.

There are fire departments that have SOPs that say any structure fire in or on a building that has solar will be fought in defensive mode only....that is only from the exterior.

 

[SolarPro]

Yes, and those concerns are codified in the International Fire Code and NFPA 1:

Designing for Fire Code Compliance (commercial rooftops)

Pitched Roof Array Layout for Fire Code Compliance (residential rooftops)

Fire Classification for Roof-Mounted PV Systems (all rooftops)

 

[Anode]

The data in the attached table are pretty interesting. Basically, SunPower—one of the largest commercial project developers in the world—found that once they closed the ground-fault detection blind spot, their PV systems stopped developing fires. They published these results here: "Resolving Fire Hazards From the Ground-Fault Detection Blind Spot."

In a related article, "The Heat Is On: Fault Detection and Fire Prevention," Bill Brooks has made a very compelling case that the industry should deploy PV systems built to NEC 2014 standards, which address both arc-fault detection and the ground-fault detection blind spot, regardless of what Code edition your AHJ enforces. This is a much more effective way to eliminate fires originating within a PV system than installing sprinkler systems under ballasted arrays.

Yes, I have heard of this extensively in lectures with Bill Brooks, Ryan Mayfield, and others over just the last few months.

To say the least, it certainly made me hyper aware of working on legacy systems we installed years ago with have the ground-fault detection blind spot. I don't want to get shocked, and certainly not worst.

 

[don_resqcapt19]

Yes, and those concerns are codified in the International Fire Code and NFPA 1:

Designing for Fire Code Compliance (commercial rooftops)

Pitched Roof Array Layout for Fire Code Compliance (residential rooftops)

Fire Classification for Roof-Mounted PV Systems (all rooftops)

Those may not always be adopted codes in the local area and the FD may have differing requirements.

 

[Anode]

Those may not always be adopted codes in the local area and the FD may have differing requirements.

Yes, and in my case, it seems they are going beyond code to the other end of the spectrum, having a little recipe of a little of this a little of that to make up the guideline...

I still can't believe I am being stonewalled so badly from this person, and that he openly said they were just, "making it up as they went along".

I mean I do understand Firefighter safety, so... But when they are unwilling to provide me the codes or clear construct to come up with a design that will exercise an exception to needing sprinklers (something they do allow, albeit made up), but will not elaborate on that without a formal plan review, is kind of ridiculous. Especially when they categorically are making certain racking systems essentially unusable, w/o the sprinkler system that is. So basically I will have to have the system engineered and incur costs, before getting answers.

I have never had issue reaching out to anyone from building, electrical, plumbing or other area, in a preliminary phase for guidance on what is legal, or how they would want something to be. No fun.

 

[SolarPro]

Those may not always be adopted codes in the local area and the FD may have differing requirements.

Not sure what your point is. Local requirements are well and good as long as they make sense at some fundamental level. Requiring sprinkler systems under a ballasted PV array in lieu of adopting consensus requirements for access pathway doesn't really fit the bill.

 

[gadfly56]

And of course there is the point that throwing water on an electrical fire is not advisable.

I can't imagine some FD engineer seriously suggesting a sprinkler system under a rooftop PV array. The roof now has to deal with the additional load of sprinkler pipe, water weight, and the reactive force when water is jetting out of the sprinklers. Due to the outside exposure it would have to be a dry system which requires an air maintenance device (compressor). It is not unknown for these compressors to fail and have the system trip and fill with water. A nice surprise when you come back from winter vacation to Washington and have to somehow replace the split pipe under your array. :thumbsdown:

 

[don_resqcapt19]

Not sure what your point is. Local requirements are well and good as long as they make sense at some fundamental level. Requiring sprinkler systems under a ballasted PV array in lieu of adopting consensus requirements for access pathway doesn't really fit the bill.

It really doesn't matter if the local requirements make sense or not...all that really matters is that the local requirements were legally adopted by the local unit of government.

It sounds like, in this case the rules have not been legally adopted and are not legally enforceable.....but you still have to get past the inspector. It may not be worth while doing the project with such restrictive requirements, and may not be worth the time and trouble to sue the inspector/AHJ for malicious prosecution.